Tile color, shape, and type detection altogether

External question — Data Science Stack Exchange Author: Mary Original post: https://datascience.stackexchange.com/questions/126637 License: CC BY-SA 4.0 — https://creativecommons.org/licenses/by-sa/4.0/ Adaptation: HTML converted to plain text; contact email addresses removed. I'm trying to apply some detections/classification on the set of tiles we have. Specifically, I need to detect color (15 classes), pattern (25 classes- on the surface of tiles, there can be certain patterns, like flower, sun, cloud, ...), and shape of the tiles (10 classes, like circle, rectangle, etc.). Tile images are under different light conditions, night/days, angles, etc. Image below is just a stock image to give you an idea of what tile is! The problem is due to some resource constraints, we can't have separate models for each and do all classifications within a single model. How can I achieve this? Any ideas?